- Email: [email protected]
Increased risk for second primary rectal cancer after pelvic radiation therapy
European Journal of Cancer 124 (2020) 142e151
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.ejcancer.com
Original Research
Increased risk for second primary rectal cancer after pelvic radiation therapy* Anouk J.M. Rombouts a,*, Niek Hugen a, Marloes A.G. Elferink b, Philip M.P. Poortmans c,d, Iris D. Nagtegaal e, Johannes H.W. de Wilt a a
Department of Surgery, Radboud University Medical Centre, Nijmegen, the Netherlands Netherlands Comprehensive Cancer Organisation, Enschede, the Netherlands c Department of Radiation Oncology, Radboud University Medical Centre, Nijmegen, the Netherlands d Department of Radiation Oncology, Institut Curie & Paris Sciences & Lettres e PSL University, Paris, France e Department of Pathology, Radboud University Medical Centre, Nijmegen, the Netherlands b
Received 27 June 2019; received in revised form 2 October 2019; accepted 15 October 2019
KEYWORDS Radiation therapy; Rectal cancer; Second cancer; Radiation-induced cancer; Prostate cancer; Endometrial cancer; Cervical cancer; Vaginal cancer; Ovarian cancer; Bladder cancer
Abstract Background: The aim of this study was to analyse the association between pelvic radiation therapy (RT) and the development of rectal cancer as a second primary cancer. Methods: Data on patients treated for a primary pelvic cancer between 1989 and 2007 were retrieved from the population-based Netherlands Cancer Registry. Patients treated for more than one pelvic cancer were excluded. To estimate the cumulative incidence of rectal cancer, Fine and Gray’s competing risk model was used with death as a competing event. Survival was calculated using multivariable Cox regression. Results: A total of 192,658 patients were included, of which 62,630 patients were treated with RT for their pelvic cancer. Primary tumours were located in the prostate (50.1%), bladder (19.2%), endometrium (13.9%), ovaries (10.0%), cervix (6.4%) and vagina (0.4%). At a median interval of 6 years (range 0e24), 1369 patients developed a rectal cancer. Overall, the risk for rectal cancer was increased in patients who underwent RT for the previous pelvic cancer (subhazard ratio [SHR]: 1.72, 95% confidence interval [CI]: 1.55e1.91). Analysis for each tumour location specifically showed an increased risk in patients who received RT for prostate (SHR: 1.89, 95% CI: 1.66e2.16) or endometrial cancer (SHR: 1.50, 95% CI: 1.13e2.00). A protective effect of RT was observed for patients with bladder cancer (SHR 0.67, 95% CI: 0.47e0.94). There was no survival difference between patients with rectal cancer with or without previous RT (hazard ratio: 0.94, 95% CI: 0.79e1.11).
*
The results of this manuscript have been presented as a poster presentation at the ESCP yearly conference of 2018 in Nice. * Corresponding author: Radboudumc, Department of Surgery, HP 618, Postal Box 9101, 6500 HB Nijmegen, the Netherlands. Fax: þ31 243635115. E-mail address: [email protected] (A.J.M. Rombouts). https://doi.org/10.1016/j.ejca.2019.10.022 0959-8049/ª 2019 Elsevier Ltd. All rights reserved.
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
143
Conclusions: Patients who received RT for a previous pelvic cancer were at increased risk for rectal cancer. The risk was modest and pronounced in patients receiving RT for prostate and endometrial cancer. ª 2019 Elsevier Ltd. All rights reserved.
1. Introduction
2. Material and methods
Compared with the general population, cancer survivors have an increased risk of developing new malignancies [1,2]. This is the result of a combination of factors such as lifestyle, genetic susceptibility and type of treatment for their previous cancer [3,4]. Radiation therapy (RT) is part of the treatment regimen in at least 50% of all patients with cancer and has been associated with the development of second primary cancers [3e5]. However, recent studies have shown that in some cases, RT does not increase the risk for a second primary cancer and might even have a preventive effect [2,6,7]. When RT is administered for a tumour located in the pelvis, the rectum is likely to be within the irradiation field and is therefore prone to early and late toxicity [8,9]. Recently, a systematic review and metaanalysis on the incidence of second primary rectal cancer after RT for a primary pelvic tumour demonstrated a small increase in rectal cancer incidence after pelvic RT [10]. However, given the sparse evidence in the literature regarding several primary tumour sites (e.g., bladder and vaginal cancer), this hypothesis warrants further study and validation using population-based data. A second primary rectal cancer after RT to the pelvic region has implications for rectal cancer treatment. One of the difficulties is whether additional courses of RT can be given because reirradiation has been associated with an excess of toxicity [11]. In addition, it is not clear whether radiation-induced rectal cancer is equally sensitive to RT compared with a sporadic rectal cancer. Yet, recent studies suggest that reirradiation can be of added value and that toxicity seems to be acceptable [12e15]. In addition, fibrosis due to previous RT can be associated with more challenging surgery and possibly increased surgical complications [15,16]. As a result, there might be a difference in outcomes between patients with rectal cancer who did or did not receive RT as part of the treatment for their previous cancer located in the pelvis. The aim of the present study was to compare the incidence of rectal cancer as a second primary cancer in patients with or without RT as treatment for their previous pelvic cancer. Moreover, data on survival of patients with rectal cancer as a second primary cancer were analysed.
Data on all patients treated for a primary cancer located in the pelvis, excluding primary rectal and anal cancer, between 1989 and 2007 were retrieved from the population-based Netherlands Cancer Registry (NCR). The NCR collects all newly diagnosed tumours. Cancers are staged according to the TNM (tumour-nodemetastasis) classification and topography, and morphology of tumours is coded using the International Classification of Diseases for Oncology (ICD-O). The NCR adheres to the ICD-O guidelines to identify multiple primary cancers and to the clinical assessment for the distinction of primary cancer from recurrent disease. The NCR performs at least annual updating on survival status. In the study period, the NCR did not differentiate between types of radiation therapy (i.e., external beam radiation therapy, brachy therapy etc.) and dosages were not registered. Tumours of the prostate, bladder, endometrium, ovaries and vagina were included as first pelvic primary cancers. Patients who were treated for more than one pelvic cancer (other than their second primary rectal cancer) and patients whose tumour was found during autopsy were excluded (N Z 3329 and N Z 72, respectively). Rectal cancers that were not classified as adenocarcinoma, mucinous adenocarcinoma or signet ring cell carcinoma were excluded (N Z 24). Information on vital status and incidence of rectal cancer as a second primary cancer was complete up to January 31st 2016 by linkage to the nationwide municipal population registries network. 2.1. Statistical analysis Statistical significance was defined as p < 0.05. SPSS statistics (version 20.0; IBM) was used for characteristics and survival analyses. Baseline characteristics were analysed using the independent sample t-test, c2 test, or in case of an expected cell count <5, the Fisher exact test. Survival outcomes were calculated using KaplaneMeier curves, and log-rank test was used for comparison of these curves. Cox regression was used for multivariable analysis, and variables were entered into the model in case of a p-value <0.1 in univariable analyses or when considered clinically relevant. Survival
144
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
time was calculated as the time between diagnosis of the rectal tumour and the date of vital status. Fine and Gray’s competing risk model was used to estimate the cumulative incidence of developing rectal cancer as a second primary cancer. The model incorporates death as a competing event for the development of a second cancer. This was performed using the Stata software (version 13.1). Subhazard ratios (SHRs) obtained from the Fine and Gray’s competing risk model describe the relative effect of covariates on the subdistribution hazard function. Follow-up was defined as the time between incidence of the first primary tumour and date of a second primary rectal tumour or the last date of study follow-up or death. Estimation of standardised incidence ratios (SIRs) was performed using SAS (version 9.2). SIRs were defined as the ratio of observed secondary rectal cancers to expected cases in the Dutch general population. Results were stratified by gender, age and calendar time; 95% confidence intervals (CIs) were calculated by Poisson regression. The absolute excess risk per 10,000 person-years was also calculated (i.e., number of observed secondary rectal cancers minus the number of expected cases divided by person-years of follow-up). 3. Results A total of 192,658 patients were included in the study, of which 62,630 patients (32.5%) were treated with RT for a primary pelvic cancer. Primary cancers were located in the prostate (50.1%), bladder (19.2%), endometrium (13.9%), ovaries (10.0%), cervix (6.4%) and vagina (0.4%; Table 1). Median follow-up was 6.8 years (range 0e27). After 10 years of follow-up, 65,992 (40.7%) patients were alive, and after 20 years, 9593 (20.9%) patients were alive. Characteristics of the previous pelvic cancers are depicted in Table 2. A total of 1369 patients (0.7%) developed rectal cancer as a second primary cancer after a median interval of 6 years (range 0e24). The crude incidence of rectal cancer was 1.0% in patients who received RT for their previous pelvic cancer and 0.6% in patients who did not receive RT (p < 0.001). Multivariable analysis (correcting for age, gender, clinical TNM stage, treatment for their previous pelvic cancer [surgery and chemotherapy], vital status, incidence year of their previous cancer and location of their previous tumour) confirmed the increased risk for rectal cancer as a second primary cancer in patients who received RT (hazard ratio [HR]: 1.19, 95% CI: 1.03e1.37). 3.1. Standardised incidence ratios When compared with the general Dutch population, patients with a history of pelvic cancer were at increased risk for the development of rectal cancer as a second primary cancer. The increased risk was caused by RT treatment, as evident by a SIR in this group of 1.20 (CI: 1.10e1.30), compared with the group of RT-naı¨ve
Table 1 Characteristics of the previous pelvic cancer. RT N Z 62,630 (%) Second primary rectal cancer Yes 618 (1.0) No 62,012 (99.0) Location of previous pelvic tumour Prostate 36,427 (58.2) Bladder 8916 (14.2) Endometrium 10,009 (16.0) Ovaries 348 (0.6) Cervix 6372 (10.2) Vagina 558 (0.9) Age, median (range) 69 (0e100) 0e14 6 (0.0) 15e29 216 (0.3) 30e44 1952 (3.1) 45e59 8871 (14.2) 60e74 37,157 (59.3) >75 14,428 (23.0) Gender Male 43,128 (68.9) Female 19,502 (31.1) Incidence year 1989e1992 9239 (14.8) 1993e1996 11,240 (17.9) 1997e2000 12,930 (20.6) 2001e2004 15,782 (25.2) 2005e2007 13,439 (21.5) Tumour differentiation Well 9829 (15.7) Intermediate 24,287 (38.8) Poor 18,366 (29.3) Undifferentiated 461 (0.7) Unknown 9687 (15.5) Treatment Surgery 19,161 (30.6) Radiation therapy 62,630 (100) Chemotherapy 2057 (3.3) Immunotherapy 50 (0.1) Hormonal therapy 13,049 (20.8)
NRT N Z 130,028 (%) 751 (0.6) 129,277 (99.4) 60,150 (46.3) 28,128 (21.6) 16,694 (12.8) 18,931 (14.6) 5983 (4.6) 142 (0.1) 69 (0e102) 101 (0.1) 902 (0.7) 6373 (4.9) 24,006 (18.5) 57,002 (43.8) 41,644 (32.0) 82,021 (63.1) 48,007 (36.9) 21,160 26,847 27,711 30,297 24,013
(16.3) (20.6) (21.3) (23.3) (18.5)
19,725 (15.2) 42,652 (32.8) 43,913 (33.8) 969 (0.7) 22,769 (17.5) 82,194 (63.2) e 23,869 (18.4) 4057 (3.1) 45,361 (34.9)
RT, radiation therapy; NRT, no radiation therapy.
patients (SIR: 0.99, CI: 0.91e1.06). The latter have a similar risk as the general Dutch population. In patients with a history of RT, the excess risk accounts for 17 cases per 100,000 patients per year. 3.2. Crude incidence ratio The crude incidence ratio of rectal cancer was increased in patients who received RT for primary prostate cancer (1.3% versus 0.7%; p < 0.001) or endometrial cancer (0.9% versus 0.6%; p Z 0.002) compared with patients who did not receive RT as part of their treatment. There was no statistically significant difference in the crude incidence ratio of rectal cancer based on previous RT for treatment of cervical (0.4% versus 0.3%, p Z 0.757), vaginal (0.4% versus 0.7%; p Z 0.494) or ovarian cancer (0.9% versus 0.3%; p Z 0.080). There was less rectal cancer in the patient group with RT-treated bladder
Table 2 Characteristics of the previous pelvic cancer divided per organ. Prostate RTþ N Z 36,427
Endometrial
Ovarian
Cervical
Vaginal
RTþ N Z 8916
RTe N Z 28,128
RTþ N Z 10,009
RTe N Z 16,694
RTþ N Z 348
RTe N Z 18,931
RTþ N Z 6372
RTe N Z 5983
RTþ N Z 558
RTe N Z 142
404 (0.7) 59,746 (99.3) 72 (2e99) 2 (<0.01) 3 (<0.01) 80 (0.1) 6856 (11.4) 29,388 (48.9) 23,821 (39.6)
39 (0.4) 8877 (99.6) 74 (1e98) 1 (<0.01) 5 (0.1) 129 (1.4) 951 (10.7) 3609 (40.5) 4221 (47.3)
180 (0.6) 27,948 (99.4) 71 (0e101) 13 (<0.01) 29 (0.1) 572 (2.0) 4507 (16.0) 13,026 (46.3) 9981 (35.5)
89 (0.9) 9920 (99.1) 67 (0e95) 1 (<0.01) 4 (<0.01) 173 (1.7) 2060 (20.6) 5310 (53.1) 2461 (24.6)
94 (0.6) 16,600 (99.4) 65 (0e102) 2 (<0.01) 19 (0.1) 499 (3.0) 5363 (32.1) 6575 (39.4) 4236 (25.4)
3 (0.9) 345 (99.1) 57 (14e94) 1 (0.3) 6 (1.7) 51 (14.7) 139 (39.9) 118 (33.9) 33 (9.5)
53 (0.3) 18,878 (99.7) 62 (0e98) 79 (0.4) 433 (2.3) 1712 (9.0) 5840 (30.8) 7464 (39.4) 3403 (18.0)
23 (0.4) 6349 (99.6) 58 (19e100) e 188 (3.0) 1536 (24.1) 1654 (26.0) 1789 (28.1) 1205 (18.9)
19 (0.3) 5694 (99.7) 40 (12e95) 2 (<0.01) 414 (6.9) 3486 (58.3) 1407 (23.5) 518 (8.7) 156 (2.6)
2 (0.4) 556 (99.6) 69 (2e95) 1 (0.2) 11 (2.0) 36 (6.5) 117 (21.0) 193 (34.6) 200 (35.8)
1 (0.7) 141 (99.3) 63 (0e95) 3 (2.1) 4 (2.8) 24 (16.9) 33 (23.2) 31 (21.8) 47 (33.1)
8340 (13.9) 12,687 (21.1) 12,835 (21.3) 14,706 (24.4) 11,582 (19.3)
2285 2071 1766 1642 1152
4627 5437 6019 6499 5546
2083 1969 2074 2129 1754
2756 3165 3504 4052 3217
(16.5) (19.0) (21.0) (24.3) (19.3)
200 (57.5) 72 (20.7) 41 (11.8) 20 (5.7) 15 (4.3)
4052 4159 4039 3922 2759
1406 1304 1353 1261 1048
(22.1) (20.5) (21.2) (19.8) (16.4)
1347 (22.5) 1369 (22.9) 1279 (21.4) 1091 (21.4) 897 (15.0)
123 (22.0) 114 (20.4) 97 (17.4) 125 (22.4) 99 (17.7)
38 30 35 27 12
146 (0.2) 9858 (16.4) 27,695 (46.0) 10,845 (18.0) 6481 (10.8) 5125(8.5)
6 (0.1) 434 (4.9) 3736 (41.9) 1825 (20.5) 1202 (13.5) 1713 (19.2)
196 (0.7) 11,386 (40.5) 5886 (20.9) 1694 (6.0) 1174 (4.2) 7792 (27.7)
66 (0.7) 4553 (45.5) 657 (6.6) 489 (4.9) 93 (0.9) 4151 (41.5)
238 (1.4) 7749 (46.4) 404 (2.4) 450 (2.7) 193 (1.2) 7660 (45.9)
4 (1.1) 177 (50.9) 67 (19.3) 44 (12.6) e 56 (16.1)
101 (0.5) 4224 (22.3) 1928 (10.2) 9892 (52.3) e 2786 (14.7)
37 (0.6) 1773 (27.8) 2438 (38.3) 1260 (19.8) 414 (6.5) 450 (7.1)
265 (4.3) 4112 (68.7) 256 (4.3) 60 (1.0) 46 (0.8) 1253 (20.9)
3 (0.5) 152 (27.2) 150 (26.9) 85 (15.2) 85 (15.2) 83 (14.9)
2 (1.4) 46 (32.4) 6 (4.2) 5 (3.5) 6 (4.2) 77 (45.2)
17,189 (29.5) 6656 (11.4) 34,395 (59.1)
5373 (60.7) 701 (7.9) 2785 (31.4)
9520 (34.0) 1513 (5.4) 16,955 (60.6)
4551 (48.3) 329 (3.5) 4543 (48.2)
6535 (42.4) 333 (2.2) 8558 (55.5)
191 (55.4) 14 (4.1) 140 (40.6)
6961 (37.0) 1988 (10.6) 9865 (52.4)
2757 (43.6) 1204 (19.0) 2365 (37.4)
2335 (39.4) 141 (2.4) 3448 (58.2)
308 (59.2) 82 (15.8) 130 (25.0)
50 (55.6) 8 (8.9) 32 (35.6)
26,551 (45.6) 18,335 (31.5) 13,355 (22.9)
7156 (80.8) 400 (4.5) 1304 (14.7)
19,719 (70.5) 1116 (4.0) 7154 (25.6)
7215 (76.6) 161 (1.7) 2047 (21.7)
10,485 (68.0) 520 (3.4) 4420 (28.7)
266 (77.1) 12 (3.5) 67 (19.4)
11,258 (59.8) 2802 (14.9) 4754 (25.3)
4735 (74.8) 343 (5.4) 1248 (19.7)
3455 (58.3) 117 (2.0) 2353 (39.7)
389 (74.8) 23 (4.4) 108 (20.8)
55 (61.1) 12 (13.3) 23 (25.6)
8285 (13.8) 23,833 (39.6) 18,198 (30.3) 242 (0.4) 9592 (15.9)
85 (1.0) 1249 (14.0) 6757 (75.8) 202 (2.3) 623 (7.0)
1368 (4.9) 8352 (29.7) 16,088 (57.2) 304 (1.1) 2016 (7.2)
2137 (21.4) 4094 (40.9) 2678 (26.8) 79 (0.8) 1021 (10.2)
7426 (44.5) 5196 (31.1) 1877 (11.2) 73 (0.4) 2122 (12.7)
30 (8.6) 123 (35.3) 85 (24.4) 9 (2.6) 101 (29.0)
2164 (11.4) 3786 (20.0) 6627 (35.0) 288 (1.5) 6066 (32.0)
285 (4.5) 2245 (35.2) 2135 (33.5) 96 (1.5) 1611 (25.3)
473 (7.9) 1457 (24.4) 1107 (18.5) 58 (1.0) 2888 (48.3)
39 (7.0) 160 (28.7) 134 (24.0) 10 (1.8) 215 (38.5)
9 (6.3) 28 (19.7) 16 (11.3) 4 (2.8) 85 (59.9)
16,825 (28.0) e 704 (1.2) 3 (<0.01) 44,342 (73.7)
6094 (68.3)
27,044 (96.1) e 8535 (30.3) 4041 (14.4) 35 (0.1)
9651 (96.4) 10009 (100) 112 (1.1) e 122 (1.2)
16,103 (96.5) e 499 (3.0) 1 (<0.01) 792 (4.7)
327 (94.0) 348 (100) 56 (16.1) 1 (0.3) 4 (1.1)
16,262 (85.9) e 13,900 (73.4) 12 (0.1) 171 (0.9)
1959 (30.7) 6372 (100) 1269 (19.9) 1 (<0.01) 19 (0.3)
5832 (97.5) e 212 (3.5) e 20 (0.3)
95 (17.0) 558 (100) 73 (13.1) e 1 (0.2)
128 (90.1) e 19 (13.4) e 1 (0.7)
RT, radiation therapy; NRT, no radiation therapy.
(25.6) (23.2) (19.8) (18.4) (12.9)
481 (5.4) 46 (0.5) 6 (0.1)
(16.4) (19.3) (21.4) (23.1) (19.7)
(20.8) (19.7) (20.7) (21.3) (17.5)
(21.4) (22.0) (21.3) (20.7) (14.6)
(26.8) (21.1) (24.6) (19.0) (8.5)
145
RTe N Z 60,150
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
Second rectal cancer Yes 462 (1.3) No 35,965 (98.7) Age, median (range) 69 (4e92) 0e14 2 (<0.01) 15e29 2 (<0.01) 30e44 27 (0.1) 45e59 3950 (10.8) 60e74 26,138 (71.8) >75 6308 (17.3) Incidence year 1989e1992 3142 (8.6) 1993e1996 5710 (15.7) 1997e2000 7599 (20.9) 2001e2004 10,605 (29.1) 2005e2008 9371 (25.7) Clinical tumour stage cT0, cTis, cTA 2 (<0.01) cT1 9363 (25.8) cT2 17,612 (48.3) cT3 8104 (22.2) cT4 838 (2.3) cTX/Unknown 508 (1.4) Clinical node stage cN0 22,299 (61.4) cNþ 649 (1.8) cNX 13,399 (36.9) Clinical metastasis stage cM0 28,798 (79.2) cM1 407 (1.1) cMX 7143 (19.7) Tumour differentiation Well 7253 (19.9) Moderate 16,416 (45.1) Poor 6577 (18.1) Undifferentiated 65 (0.2) Unknown 6116 (16.8) Treatment Surgery 1035 (2.8) Radiation therapy 36427 (100) Chemotherapy 66 (0.2) Immunotherapy 2 (<0.01) Hormonal therapy 12,987 (35.4)
Bladder
146
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
cancer compared with patients with RT-naı¨ve bladder cancer (crude incidence 0.4% versus 0.6%; p Z 0.018). 3.3. Competing risk analysis Competing risk analysis was performed to correct for death as a competing event (see Fig. 1). For all first tumour locations taken together, RT increased the risk for rectal cancer as a second primary cancer (SHR: 1.72, 95% CI: 1.55e1.91). Age did not influence this relation.
Competing risks were also calculated for each primary tumour location specifically. The risk for rectal cancer was increased in patients who received RT for prostate cancer (SHR: 1.89, 95% CI: 1.66e2.16) or endometrial cancer (SHR: 1.50, 95% CI: 1.13e2.00). Yet, a protective effect of RT was observed for patients who received RT for bladder cancer (SHR: 0.67, 95% CI: 0.47e0.94). There was no interaction with gender. No effect was observed in patients who received RT for cervical cancer (SHR: 1.17, 95% CI: 0.64e2.15), ovarian cancer (SHR:
Fig. 1. Cumulative incidence of second primary rectal cancer RT after RT for (A) any pelvic cancer; (B) prostate cancer; (C) cervical cancer; (D) endometrial cancer; (E) ovarian cancer; (F) vaginal cancer and (G) bladder cancer. RT radiation therapy; NRT, no radiation therapy.
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
2.78, 95% CI: 0.86e8.78) or vaginal cancer (SHR: 0.52, 95% CI: 0.05e5.88). 3.4. Characteristics of rectal cancer and survival analysis The median age at the time of diagnosis of rectal cancer was 75 years (range, 44e97), and most patients with a Table 3 Characteristics of rectal cancer as a second primary cancer.
Age, median (range) Gender Male Female Incidence year 1989e1992 1993e1996 1997e2000 2001e2004 2005e2008 Histology adenocarcinoma mucinous carcinoma signet ring cell carcinoma Tumour differentiation well moderate poor undifferentiated not applicable/ unknown Treatment Sx Sx þ CT Sx þ RT Sx þ CRT CT RT CRT No Sx or (C)RT Clinical tumour stage cT0/cTis cT1 cT2 cT3 cT4 cTx/unknown Clinical node stage cN0 cNþ cNx/unknown Clinical metastasis stage cM0 cM1 cMx/unknown
RT for previous pelvic cancer N Z 618 (%)
NRT for previous ppelvic cancer value N Z 751 (%)
75 (45e97)
75 (44e95)
497 (80.4) 121 (19.6)
562 (74.8) 189 (25.2)
10 (3.0) 31 (9.2) 53 (15.8) 90 (26.8) 152 (45.2)
16 (3.4) 48 (10.1) 90 (19.0) 133 (28.1) 187 (39.5)
566 (91.6) 46 (7.4)
675 (89.9) 69 (9.2)
6 (1.0)
7 (0.9)
0.002 0.014
0.541
0.511
147
rectal cancer were men (77.4%). Clinical TNM stage, differentiation grade and histology were similar between the groups (Table 3). Patients who received RT for their previous pelvic cancer were less likely to undergo additional RT for the rectal cancer (15.6% versus 46.8%, respectively, p < 0.001). Subsequently, patients who were treated with RT for the primary pelvic cancer were diagnosed with higher pathological T-stage of the rectal cancer (Table 4). There were no differences in pathological nodal stage or metastasis stage for rectal cancer in patients treated with or without RT for the primary pelvic cancer. Univariable analysis showed no statistically significant difference in overall survival between patients with rectal cancer based on their history of RT: 5-year overall survival after diagnosis of rectal cancer was 33.7% (95% CI: 29.6e37.8) in patients treated with RT for their previous pelvic cancer versus 39.1% (95% CI: 35.4e42.8) in patients without RT (Fig. 2). Multivariable analysis, correcting for age, gender, incidence year, stage and treatment of the rectal cancer, confirmed this (HR: 0.94, 95% CI: 0.79e1.11).
4. Discussion 0.182 26 (5.7) 359 (78.7) 67 (14.7) 4 (0.9) 162
31 (6.4) 357 (73.8) 94 (19.4) 2 (0.4) 267
365 (59.1) 27 (4.4) 42 (6.8) 33 (5.3) 24 (3.9) 12 (1.9) 10 (1.6) 105 (17.0)
258 (34.4) 23 (3.1) 221 (29.4) 70 (9.3) 24 (3.2) 48 (6.4) 15 (2.0) 92 (12.3)
<0.001
0.797 1 (0.3) 20 (6.1) 79 (24.0) 149 (45.3) 80 (24.3) 272
2 (0.6) 23 (6.7) 77 (22.5) 167 (48.8) 73 (21.3) 394
259 (70.4) 109 (29.6) 233
274 (65.1) 147 (34.9) 314
0.113
0.881 449 (82.2) 97 (17.8) 55
521 (82.6) 110 (17.4) 105
RT, radiation therapy; NRT, no radiation therapy; Sx, surgery; CT, chemotherapy; CRT, chemoradiation.
A recent systematic review and metaanalysis showed an increased risk of rectal cancer after RT for a variety of pelvic cancers [10]. Stratification for origin of the first cancer showed that the risk for rectal cancer was increased after treatment for prostate, cervical or endometrial cancer. The present nationwide study is the largest patient cohort in the literature and analysed the association between RT as treatment for pelvic cancer and the development of rectal cancer as a second primary cancer. The current data confirms that patients who receive RT for a primary pelvic cancer are at increased risk for the development of rectal cancer as a second primary cancer. Analysis of each primary cancer location specifically showed an increased risk for patients who received RT for a primary prostate or endometrial cancer. In contrast, a small preventive effect of RT on rectal cancer development was seen for patients who received RT for bladder cancer and no influence of RT was observed for patients who received RT for the other gynaecological cancers. Little is known about the aetiology of rectal cancer in relation to a history of RT. RT effects can be divided into set and random effects according to the probability of occurrence [17]. Set effects, such as teratogenesis, occur above a threshold dose and the severity increases with increasing the dose. In contrast, the probability for a random effect such as carcinogenesis increases as the dose increases. In that case, there is no minimal threshold dose, and the severity is independent of the dose. Since the introduction of intensity-modulated radiation therapy (IMRT), and even more volumetric
148
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
Table 4 Pathological tumour stage of rectal cancer as a second primary cancer.
No (chemo)radiation for rectal cancer
No (chemo)radiation for rectal cancer
Pathological tumour stage pT1 pT2 pT3 pT4 pTx/unknown Pathological nodal stage pN0 pNþ pNx/unknown Metastasis stage M0 M1 Mx/unknown
Pathological tumour stage ypT0 ypT1 ypT2 ypT3 ypT4 ypTx/unknown Pathological nodal stage ypN0 ypNþ ypNx/unknown Metastasis stage M0 M1 Mx/unknown
RT for previous pelvic cancer N Z 382 (%)
NRT for previous pelvic cancer N Z 277 (%)
40 (11.0) 115 (31.6) 186 (51.1) 23 (6.3) 18
69 (26.4) 82 (31.4) 100 (38.3) 10 (3.8) 16
179 (61.5) 112 (38.5) 91
107 (61.1) 68 (38.9) 102
319 (93.0) 24 (7.0) 39
213 (92.6) 17 (7.4) 47
RT for previous pelvic cancer N Z 75 (%)
NRT for previous pelvic cancer N Z 291 (%)
5 (6.9) 1 (1.4) 23 (31.9) 34 (47.2) 9 (12.5) 3
11 (4.0) 30 (10.8) 87 (31.3) 142 (51.1) 8 (2.9) 13
46 (69.7) 20 (30.3) 9
150 (58.8) 105 (41.2) 36
62 (91.2) 6 (8.8) 7
254 (93.7) 17 (6.3) 20
p-value <0.001
0.937
0.858
p-value 0.001
0.106
0.427
Pathological tumour stage of patients who underwent surgery for rectal cancer, with or without (chemo)radiation. Fourteen patients underwent neoadjuvant chemotherapy without radiation and are not taken into consideration in this table. RT, radiation therapy; NRT, no radiation therapy.
IMRT, where the high-dose region is geometrically more shaped around the target volumes at a cost of a socalled low-dose bath to a larger volume of non-target
Fig. 2. KaplaneMeier curve for overall survival after diagnosis of rectal cancer. RT radiation therapy for previous pelvic cancer; NRT, no radiation therapy for previous pelvic cancer.
tissue, it was argued that these differences in RT dose distribution might increase the risk for induction of second cancers [10,18e20]. However, little data are available to support this. There are several other possible explanations for the observed discrepancies in rectal cancer risks between the primary irradiated organs. Firstly, genetic susceptibility is likely to play a role. For instance, patients with endometrial cancer with Lynch syndrome have a 50% chance of developing a subsequent colorectal cancer [21]. In addition, DNA mismatch repair (MMR) status has been associated with radio sensitivity and long-term toxicity [22]. Increase in MMR expression is related to low-dose rate sensitivity, whereas MMR deficiency is related to high-dose rate sensitivity and possibly increased risk of carcinogenesis [22]. Secondly, because of its fixed location in the pelvis, the rectum is prone for toxicity after RT, and chronic proctitis occurs in 2e20% of patients receiving RT to the pelvis [23]. Hypothetically, RT-induced late toxicity leading to inflammation could create a precancerous environment, as is observed in patients with inflammatory bowel diseases (IBD) [24e26]. Carcinogenesis of IBD-related colorectal cancer differs from sporadic
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
colorectal cancer [25,26]. Sporadic colorectal cancer typically develops via the adenoma-carcinoma sequence, characterised by a series of molecular alterations (e.g., RAS mutation or p53 mutation). In IBD-related colorectal cancer, the ‘inflammation-dysplasia-carcinoma’ sequence shows a different pattern of molecular alterations [27]. Carcinogenesis in patients with IBD is characterised by a rapid development of high-grade dysplasia. Because RT is associated with chronic proctitis, we hypothesise that the development of rectal cancer after RT might be similar to colorectal cancer in patients with IBD. If that is the case, the incidence of rectal cancer would be related to toxicity experienced by patients after RT. For example, late grade 2 toxicity (following the Radiation Therapy Oncology Group [RTOG] and the European Organization for Research and Treatment of Cancer [EORTC]. radiation morbidity scoring schema) occurs in approximately 3e7% of patients with bladder cancer, 5e20% in patients with prostate cancer, 10e20% in patients with cervical cancer and up to 22% in patients with endometrial cancer [28e33]. In combination with the findings in the present study, it could be hypothesised that this may be an explanation for the increased risk of rectal cancer in patients with prostate and endometrial cancer. Another reason of the higher incidence of rectal cancer in patients with prostate cancer might be the dose/ volume characteristics of up to 70Gy to a small part of the anterior rectal wall in prostate cancer versus 45e50Gy to much larger volumes for most other pelvic tumours. On the other hand, numbers between groups varied and observed risk differences were small. Lack of power in the smaller groups might be one of the reasons for the lack of effect of RT for cervical, vaginal and ovarian cancer. Theoretically, differences in RT techniques, that is, external beam radiation versus brachytherapy might also play a role, but this could not be confirmed in our recent review of the literature [10]. In the NCR, no data were available on RT techniques used. In the earlier years most patients in the Netherlands received pelvic RT according to the traditional ‘box’ technique. During the 1990s, 3D conformal computed tomography (CT)e based techniques were gradually introduced, enabling individualisation to volume- instead of field-based RT. However, during multivariable analysis, incidence year of the primary cancer was not a significant predictor for the development of a rectal cancer. Differences in predisposing lifestyle factors such as smoking might also play a role in the development of second cancers. Hegemann et al. [34] have studied the incidence of second cancers after treatment for prostate cancer and concluded that the lack of information on smoking is one of the major limitations in the available literature. They reported on an increased risk for second rectal cancer in patients who were treated with RT only, but no increase in second rectal cancer was found in patients treated with surgery followed by RT or surgery
149
only. The authors hypothesised that the difference in second rectal cancer incidence might be due to differences in age and lifestyle habits such as smoking. In the present study, 36,427 patients with prostate cancer who received RT and 60,150 RT-naı¨ve patients were analysed. In contrast to the study by Hegemann et al. [34], age was not a significant predictor for the incidence of second rectal cancer. Unfortunately, information on smoking habits is not available in the NCR, and therefore, a possible confounding effect could not be studied. There was no significant difference in survival after the diagnosis of rectal cancer depending on the patients’ history of RT. Clinical stage of the second primary rectal cancers was comparable and perhaps even favourable to clinical stage reported in the literature for primary rectal cancer [35,36]. Notwithstanding, survival rates were lower than those reported for the general population (34e39% versus 51e65%) [36,37]. Patients with rectal cancer in the present study were slightly older than patients with a sporadic rectal cancer described in the literature [35,36]. Moreover, the fact that these patients have a history of cancer and cancer treatment, which might have affected their fitness and outcome, could explain the difference in overall survival. Another hypothesis could be that, tumours of patients who present with a second primary cancer might have different biological behaviour and/or might be more aggressive than tumours in patients who present with a first primary cancer [38]. This is supported by the advanced T-stage of patients with a history of RT and the relative high number of ypT4 cases after reirradiation in the present study. This nationwide study provides unique long-term data on the incidence and outcomes of second primary rectal cancer after RT for primary pelvic cancer. One of the strengths of our study is that, it contains a large cohort of patients with more than 65,000 patients having more than 10 years of follow-up and 9500 patients more than 20 years of follow-up. However, the nature of the study needs to be taken into account when interpreting the data. To diminish the influence of confounding factors, multivariable analysis was conducted. However, for some of the known predisposing factors, such as genetic susceptibility and lifestyle factors, no data were available, which is considered a limitation of the study. Only few patients underwent RT for a primary ovarian or vaginal cancer, and therefore, we could not establish whether RT for the primary cancer was related to an increased risk of second rectal cancer. However, because of the low number of patients with ovarian and vaginal cancer receiving RT, a possible relation between RT and second primary rectal cancer would not be clinically relevant.
5. Conclusions Patients who received RT for pelvic cancer were at increased risk for rectal cancer as a second primary
150
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
cancer. However, its incidence is low and analysis per primary pelvic organ specifically showed an increased risk after prostate and endometrial cancer treatment only. Patients with previous RT for bladder cancer had a decreased risk for rectal cancer. There was no difference in survival after diagnosis of rectal cancer depending on their history of RT.
[13]
[14]
[15]
Conflict of interest statement There are no conflicts of interest to declare. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
[16]
[17]
References [1] Marijnen CA, Nagtegaal ID, Klein Kranenbarg E, Hermans J, van de Velde CJ, Leer JW, et al. No downstaging after short-term preoperative radiotherapy in rectal cancer patients. J Clin Oncol 2001;19:1976e84. [2] Rombouts AJM, Hugen N, Elferink MAG, Feuth T, Poortmans PMP, Nagtegaal ID, et al. Incidence of second tumors after treatment with or without radiation for rectal cancer. Ann Oncol 2017;28:535e40. [3] Berrington de Gonzalez A, Curtis RE, Kry SF, Gilbert E, Lamart S, Berg CD, et al. Proportion of second cancers attributable to radiotherapy treatment in adults: a cohort study in the US SEER cancer registries. Lancet Oncol 2011;12:353e60. [4] Birgisson H, Pahlman L, Gunnarsson U, Glimelius B. Occurrence of second cancers in patients treated with radiotherapy for rectal cancer. J Clin Oncol 2005;23:6126e31. [5] Rombouts AJM, Huising J, Hugen N, Siesling S, Poortmans PM, Nagtegaal ID, et al. Assessment of radiotherapy-associated angiosarcoma after breast cancer treatment in a Dutch population-based study. JAMA Oncol 2019;5(2):267e9. [6] Warschkow R, Guller U, Cerny T, Schmied BM, Plasswilm L, Putora PM. Secondary malignancies after rectal cancer resection with and without radiation therapy: a propensity-adjusted, population-based SEER analysis. Radiother Oncol 2017;123:139e46. [7] Wiltink LM, Nout RA, Fiocco M, Meershoek-Klein Kranenbarg E, Jurgenliemk-Schulz IM, Jobsen JJ, et al. No increased risk of second cancer after radiotherapy in patients treated for rectal or endometrial cancer in the randomized TME, PORTEC-1, and PORTEC-2 Trials. J Clin Oncol 2015;33: 1640e6. [8] Stenmark MH, Conlon AS, Johnson S, Daignault S, Litzenberg D, Marsh R, et al. Dose to the inferior rectum is strongly associated with patient reported bowel quality of life after radiation therapy for prostate cancer. Radiother Oncol 2014; 110:291e7. [9] Harkenrider MM, Alite F, Silva SR, Small Jr W. Image-based brachytherapy for the treatment of cervical cancer. Int J Radiat Oncol Biol Phys 2015;92:921e34. [10] Rombouts AJM, Hugen N, van Beek JJP, Poortmans PMP, de Wilt JHW, Nagtegaal ID. Does pelvic radiation increase rectal cancer incidence? a systematic review and meta-analysis. Cancer Treat Rev 2018;68:136e44. [11] Glimelius B. Recurrent rectal cancer. The pre-irradiated primary tumour: can more radiotherapy be given? Colorectal Dis 2003;5: 501e3. [12] van der Meij W, Rombouts AJ, Rutten H, Bremers AJ, de Wilt JH. Treatment of locally recurrent rectal carcinoma in
[18]
[19]
[20]
[21]
[22]
[23]
[24] [25]
[26]
[27]
[28]
[29]
[30]
previously (chemo)irradiated patients: a review. Dis Colon Rectum 2016;59:148e56. Susko M, Lee J, Salama J, Thomas S, Uronis H, Hsu D, et al. The use of re-irradiation in locally recurrent, non-metastatic rectal cancer. Ann Surg Oncol 2016;23:3609e15. Guren MG, Undseth C, Rekstad BL, Braendengen M, Dueland S, Spindler KL, et al. Reirradiation of locally recurrent rectal cancer: a systematic review. Radiother Oncol 2014;113: 151e7. Valentini V, Morganti AG, Gambacorta MA, Mohiuddin M, Doglietto GB, Coco C, et al. Preoperative hyperfractionated chemoradiation for locally recurrent rectal cancer in patients previously irradiated to the pelvis: a multicentric phase II study. Int J Radiat Oncol Biol Phys 2006;64:1129e39. Rombouts AJ, Koh CE, Young JM, Masya L, Roberts R, DeLoyde K, et al. Does radiotherapy of the primary rectal cancer affect prognosis after pelvic exenteration for recurrent rectal cancer? Dis Colon Rectum 2015;58:65e73. ICRP. ICoRP. Publication 84. Pregnancy and medical irradiation. Ann ICRP 2000;30:1e43. Kry SF, Salehpour M, Followill DS, Stovall M, Kuban DA, White RA, et al. The calculated risk of fatal secondary malignancies from intensity-modulated radiation therapy. Int J Radiat Oncol Biol Phys 2005;62:1195e203. Berrington de Gonzalez A, Wong J, Kleinerman R, Kim C, Morton L, Bekelman JE. Risk of second cancers according to radiation therapy technique and modality in prostate cancer survivors. Int J Radiat Oncol Biol Phys 2015;91:295e302. Lonn S, Gilbert ES, Ron E, Smith SA, Stovall M, Curtis RE. Comparison of second cancer risks from brachytherapy and external beam therapy after uterine corpus cancer. Cancer Epidemiol Biomark Prev 2010;19:464e74. Win AK, Lindor NM, Winship I, Tucker KM, Buchanan DD, Young JP, et al. Risks of colorectal and other cancers after endometrial cancer for women with Lynch syndrome. J Natl Cancer Inst 2013;105:274e9. Martin LM, Marples B, Coffey M, Lawler M, Lynch TH, Hollywood D, et al. DNA mismatch repair and the DNA damage response to ionizing radiation: making sense of apparently conflicting data. Cancer Treat Rev 2010;36:518e27. Wu XR, Liu XL, Katz S, Shen B. Pathogenesis, diagnosis, and management of ulcerative proctitis, chronic radiation proctopathy, and diversion proctitis. Inflamm Bowel Dis 2015;21: 703e15. Sjodahl RI, Myrelid P, Soderholm JD. Anal and rectal cancer in Crohn’s disease. Colorectal Dis 2003;5:490e5. Dyson JK, Rutter MD. Colorectal cancer in inflammatory bowel disease: what is the real magnitude of the risk? World J Gastroenterol 2012;18:3839e48. Baker KT, Salk JJ, Brentnall TA, Risques RA. Precancer in ulcerative colitis: the role of the field effect and its clinical implications. Carcinogenesis 2018;39:11e20. Keller DS, Windsor A, Cohen R, Chand M. Colorectal cancer in inflammatory bowel disease: review of the evidence. Tech Coloproctol 2019;23:3e13. Aluwini S, van Rooij PH, Kirkels WJ, Boormans JL, KolkmanDeurloo IK, Wijnmaalen A. Bladder function preservation with brachytherapy, external beam radiation therapy, and limited surgery in bladder cancer patients: long-term results [corrected]. Int J Radiat Oncol Biol Phys 2014;88:611e7. Chen CC, Wang L, Lu CH, Lin JC, Jan JS. Comparison of clinical outcomes and toxicity in endometrial cancer patients treated with adjuvant intensity-modulated radiation therapy or conventional radiotherapy. J Formos Med Assoc 2014;113: 949e55. Kishan AU, Kupelian PA. Late rectal toxicity after low-dose-rate brachytherapy: incidence, predictors, and management of side effects. Brachytherapy 2015;14:148e59.
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151 [31] Mohammed N, Kestin L, Ghilezan M, Krauss D, Vicini F, Brabbins D, et al. Comparison of acute and late toxicities for three modern high-dose radiation treatment techniques for localized prostate cancer. Int J Radiat Oncol Biol Phys 2012;82: 204e12. [32] Chopra S, Dora T, Engineer R, Mechanery S, Agarwal P, Kannan S, et al. Late rectal toxicity after image-based high-doserate interstitial brachytherapy for postoperative recurrent and/or residual cervical cancers: EQD2 predictors for Grade >/ZII toxicity. Brachytherapy 2015;14:881e8. [33] Mattes MD, Kan CC, Dalbagni G, Zelefsky MJ, Kollmeier MA. External beam radiation therapy for small cell carcinoma of the urinary bladder. Pract Radiat Oncol 2015;5:e17e22. [34] Hegemann NS, Schlesinger-Raab A, Ganswindt U, Horl C, Combs SE, Holzel D, et al. Risk of second cancer following radiotherapy for prostate cancer: a population-based analysis. Radiat Oncol 2017;12:2.
151
[35] Benitez Majano S, Di Girolamo C, Rachet B, Maringe C, Guren MG, Glimelius B, et al. Surgical treatment and survival from colorectal cancer in Denmark, England, Norway, and Sweden: a population-based study. Lancet Oncol 2019;20:74e87. [36] Brouwer NPM, Bos A, Lemmens V, Tanis PJ, Hugen N, Nagtegaal ID, et al. An overview of 25 years of incidence, treatment and outcome of colorectal cancer patients. Int J Cancer 2018;143:2758e66. [37] Peeters KC, Marijnen CA, Nagtegaal ID, Kranenbarg EK, Putter H, Wiggers T, et al. The TME trial after a median followup of 6 years: increased local control but no survival benefit in irradiated patients with resectable rectal carcinoma. Ann Surg 2007;246:693e701. [38] VanderWalde AM, Hurria A. Second malignancies among elderly survivors of cancer. The Oncologist 2011;16:1572e81.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.ejcancer.com
Original Research
Increased risk for second primary rectal cancer after pelvic radiation therapy* Anouk J.M. Rombouts a,*, Niek Hugen a, Marloes A.G. Elferink b, Philip M.P. Poortmans c,d, Iris D. Nagtegaal e, Johannes H.W. de Wilt a a
Department of Surgery, Radboud University Medical Centre, Nijmegen, the Netherlands Netherlands Comprehensive Cancer Organisation, Enschede, the Netherlands c Department of Radiation Oncology, Radboud University Medical Centre, Nijmegen, the Netherlands d Department of Radiation Oncology, Institut Curie & Paris Sciences & Lettres e PSL University, Paris, France e Department of Pathology, Radboud University Medical Centre, Nijmegen, the Netherlands b
Received 27 June 2019; received in revised form 2 October 2019; accepted 15 October 2019
KEYWORDS Radiation therapy; Rectal cancer; Second cancer; Radiation-induced cancer; Prostate cancer; Endometrial cancer; Cervical cancer; Vaginal cancer; Ovarian cancer; Bladder cancer
Abstract Background: The aim of this study was to analyse the association between pelvic radiation therapy (RT) and the development of rectal cancer as a second primary cancer. Methods: Data on patients treated for a primary pelvic cancer between 1989 and 2007 were retrieved from the population-based Netherlands Cancer Registry. Patients treated for more than one pelvic cancer were excluded. To estimate the cumulative incidence of rectal cancer, Fine and Gray’s competing risk model was used with death as a competing event. Survival was calculated using multivariable Cox regression. Results: A total of 192,658 patients were included, of which 62,630 patients were treated with RT for their pelvic cancer. Primary tumours were located in the prostate (50.1%), bladder (19.2%), endometrium (13.9%), ovaries (10.0%), cervix (6.4%) and vagina (0.4%). At a median interval of 6 years (range 0e24), 1369 patients developed a rectal cancer. Overall, the risk for rectal cancer was increased in patients who underwent RT for the previous pelvic cancer (subhazard ratio [SHR]: 1.72, 95% confidence interval [CI]: 1.55e1.91). Analysis for each tumour location specifically showed an increased risk in patients who received RT for prostate (SHR: 1.89, 95% CI: 1.66e2.16) or endometrial cancer (SHR: 1.50, 95% CI: 1.13e2.00). A protective effect of RT was observed for patients with bladder cancer (SHR 0.67, 95% CI: 0.47e0.94). There was no survival difference between patients with rectal cancer with or without previous RT (hazard ratio: 0.94, 95% CI: 0.79e1.11).
*
The results of this manuscript have been presented as a poster presentation at the ESCP yearly conference of 2018 in Nice. * Corresponding author: Radboudumc, Department of Surgery, HP 618, Postal Box 9101, 6500 HB Nijmegen, the Netherlands. Fax: þ31 243635115. E-mail address: [email protected] (A.J.M. Rombouts). https://doi.org/10.1016/j.ejca.2019.10.022 0959-8049/ª 2019 Elsevier Ltd. All rights reserved.
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
143
Conclusions: Patients who received RT for a previous pelvic cancer were at increased risk for rectal cancer. The risk was modest and pronounced in patients receiving RT for prostate and endometrial cancer. ª 2019 Elsevier Ltd. All rights reserved.
1. Introduction
2. Material and methods
Compared with the general population, cancer survivors have an increased risk of developing new malignancies [1,2]. This is the result of a combination of factors such as lifestyle, genetic susceptibility and type of treatment for their previous cancer [3,4]. Radiation therapy (RT) is part of the treatment regimen in at least 50% of all patients with cancer and has been associated with the development of second primary cancers [3e5]. However, recent studies have shown that in some cases, RT does not increase the risk for a second primary cancer and might even have a preventive effect [2,6,7]. When RT is administered for a tumour located in the pelvis, the rectum is likely to be within the irradiation field and is therefore prone to early and late toxicity [8,9]. Recently, a systematic review and metaanalysis on the incidence of second primary rectal cancer after RT for a primary pelvic tumour demonstrated a small increase in rectal cancer incidence after pelvic RT [10]. However, given the sparse evidence in the literature regarding several primary tumour sites (e.g., bladder and vaginal cancer), this hypothesis warrants further study and validation using population-based data. A second primary rectal cancer after RT to the pelvic region has implications for rectal cancer treatment. One of the difficulties is whether additional courses of RT can be given because reirradiation has been associated with an excess of toxicity [11]. In addition, it is not clear whether radiation-induced rectal cancer is equally sensitive to RT compared with a sporadic rectal cancer. Yet, recent studies suggest that reirradiation can be of added value and that toxicity seems to be acceptable [12e15]. In addition, fibrosis due to previous RT can be associated with more challenging surgery and possibly increased surgical complications [15,16]. As a result, there might be a difference in outcomes between patients with rectal cancer who did or did not receive RT as part of the treatment for their previous cancer located in the pelvis. The aim of the present study was to compare the incidence of rectal cancer as a second primary cancer in patients with or without RT as treatment for their previous pelvic cancer. Moreover, data on survival of patients with rectal cancer as a second primary cancer were analysed.
Data on all patients treated for a primary cancer located in the pelvis, excluding primary rectal and anal cancer, between 1989 and 2007 were retrieved from the population-based Netherlands Cancer Registry (NCR). The NCR collects all newly diagnosed tumours. Cancers are staged according to the TNM (tumour-nodemetastasis) classification and topography, and morphology of tumours is coded using the International Classification of Diseases for Oncology (ICD-O). The NCR adheres to the ICD-O guidelines to identify multiple primary cancers and to the clinical assessment for the distinction of primary cancer from recurrent disease. The NCR performs at least annual updating on survival status. In the study period, the NCR did not differentiate between types of radiation therapy (i.e., external beam radiation therapy, brachy therapy etc.) and dosages were not registered. Tumours of the prostate, bladder, endometrium, ovaries and vagina were included as first pelvic primary cancers. Patients who were treated for more than one pelvic cancer (other than their second primary rectal cancer) and patients whose tumour was found during autopsy were excluded (N Z 3329 and N Z 72, respectively). Rectal cancers that were not classified as adenocarcinoma, mucinous adenocarcinoma or signet ring cell carcinoma were excluded (N Z 24). Information on vital status and incidence of rectal cancer as a second primary cancer was complete up to January 31st 2016 by linkage to the nationwide municipal population registries network. 2.1. Statistical analysis Statistical significance was defined as p < 0.05. SPSS statistics (version 20.0; IBM) was used for characteristics and survival analyses. Baseline characteristics were analysed using the independent sample t-test, c2 test, or in case of an expected cell count <5, the Fisher exact test. Survival outcomes were calculated using KaplaneMeier curves, and log-rank test was used for comparison of these curves. Cox regression was used for multivariable analysis, and variables were entered into the model in case of a p-value <0.1 in univariable analyses or when considered clinically relevant. Survival
144
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
time was calculated as the time between diagnosis of the rectal tumour and the date of vital status. Fine and Gray’s competing risk model was used to estimate the cumulative incidence of developing rectal cancer as a second primary cancer. The model incorporates death as a competing event for the development of a second cancer. This was performed using the Stata software (version 13.1). Subhazard ratios (SHRs) obtained from the Fine and Gray’s competing risk model describe the relative effect of covariates on the subdistribution hazard function. Follow-up was defined as the time between incidence of the first primary tumour and date of a second primary rectal tumour or the last date of study follow-up or death. Estimation of standardised incidence ratios (SIRs) was performed using SAS (version 9.2). SIRs were defined as the ratio of observed secondary rectal cancers to expected cases in the Dutch general population. Results were stratified by gender, age and calendar time; 95% confidence intervals (CIs) were calculated by Poisson regression. The absolute excess risk per 10,000 person-years was also calculated (i.e., number of observed secondary rectal cancers minus the number of expected cases divided by person-years of follow-up). 3. Results A total of 192,658 patients were included in the study, of which 62,630 patients (32.5%) were treated with RT for a primary pelvic cancer. Primary cancers were located in the prostate (50.1%), bladder (19.2%), endometrium (13.9%), ovaries (10.0%), cervix (6.4%) and vagina (0.4%; Table 1). Median follow-up was 6.8 years (range 0e27). After 10 years of follow-up, 65,992 (40.7%) patients were alive, and after 20 years, 9593 (20.9%) patients were alive. Characteristics of the previous pelvic cancers are depicted in Table 2. A total of 1369 patients (0.7%) developed rectal cancer as a second primary cancer after a median interval of 6 years (range 0e24). The crude incidence of rectal cancer was 1.0% in patients who received RT for their previous pelvic cancer and 0.6% in patients who did not receive RT (p < 0.001). Multivariable analysis (correcting for age, gender, clinical TNM stage, treatment for their previous pelvic cancer [surgery and chemotherapy], vital status, incidence year of their previous cancer and location of their previous tumour) confirmed the increased risk for rectal cancer as a second primary cancer in patients who received RT (hazard ratio [HR]: 1.19, 95% CI: 1.03e1.37). 3.1. Standardised incidence ratios When compared with the general Dutch population, patients with a history of pelvic cancer were at increased risk for the development of rectal cancer as a second primary cancer. The increased risk was caused by RT treatment, as evident by a SIR in this group of 1.20 (CI: 1.10e1.30), compared with the group of RT-naı¨ve
Table 1 Characteristics of the previous pelvic cancer. RT N Z 62,630 (%) Second primary rectal cancer Yes 618 (1.0) No 62,012 (99.0) Location of previous pelvic tumour Prostate 36,427 (58.2) Bladder 8916 (14.2) Endometrium 10,009 (16.0) Ovaries 348 (0.6) Cervix 6372 (10.2) Vagina 558 (0.9) Age, median (range) 69 (0e100) 0e14 6 (0.0) 15e29 216 (0.3) 30e44 1952 (3.1) 45e59 8871 (14.2) 60e74 37,157 (59.3) >75 14,428 (23.0) Gender Male 43,128 (68.9) Female 19,502 (31.1) Incidence year 1989e1992 9239 (14.8) 1993e1996 11,240 (17.9) 1997e2000 12,930 (20.6) 2001e2004 15,782 (25.2) 2005e2007 13,439 (21.5) Tumour differentiation Well 9829 (15.7) Intermediate 24,287 (38.8) Poor 18,366 (29.3) Undifferentiated 461 (0.7) Unknown 9687 (15.5) Treatment Surgery 19,161 (30.6) Radiation therapy 62,630 (100) Chemotherapy 2057 (3.3) Immunotherapy 50 (0.1) Hormonal therapy 13,049 (20.8)
NRT N Z 130,028 (%) 751 (0.6) 129,277 (99.4) 60,150 (46.3) 28,128 (21.6) 16,694 (12.8) 18,931 (14.6) 5983 (4.6) 142 (0.1) 69 (0e102) 101 (0.1) 902 (0.7) 6373 (4.9) 24,006 (18.5) 57,002 (43.8) 41,644 (32.0) 82,021 (63.1) 48,007 (36.9) 21,160 26,847 27,711 30,297 24,013
(16.3) (20.6) (21.3) (23.3) (18.5)
19,725 (15.2) 42,652 (32.8) 43,913 (33.8) 969 (0.7) 22,769 (17.5) 82,194 (63.2) e 23,869 (18.4) 4057 (3.1) 45,361 (34.9)
RT, radiation therapy; NRT, no radiation therapy.
patients (SIR: 0.99, CI: 0.91e1.06). The latter have a similar risk as the general Dutch population. In patients with a history of RT, the excess risk accounts for 17 cases per 100,000 patients per year. 3.2. Crude incidence ratio The crude incidence ratio of rectal cancer was increased in patients who received RT for primary prostate cancer (1.3% versus 0.7%; p < 0.001) or endometrial cancer (0.9% versus 0.6%; p Z 0.002) compared with patients who did not receive RT as part of their treatment. There was no statistically significant difference in the crude incidence ratio of rectal cancer based on previous RT for treatment of cervical (0.4% versus 0.3%, p Z 0.757), vaginal (0.4% versus 0.7%; p Z 0.494) or ovarian cancer (0.9% versus 0.3%; p Z 0.080). There was less rectal cancer in the patient group with RT-treated bladder
Table 2 Characteristics of the previous pelvic cancer divided per organ. Prostate RTþ N Z 36,427
Endometrial
Ovarian
Cervical
Vaginal
RTþ N Z 8916
RTe N Z 28,128
RTþ N Z 10,009
RTe N Z 16,694
RTþ N Z 348
RTe N Z 18,931
RTþ N Z 6372
RTe N Z 5983
RTþ N Z 558
RTe N Z 142
404 (0.7) 59,746 (99.3) 72 (2e99) 2 (<0.01) 3 (<0.01) 80 (0.1) 6856 (11.4) 29,388 (48.9) 23,821 (39.6)
39 (0.4) 8877 (99.6) 74 (1e98) 1 (<0.01) 5 (0.1) 129 (1.4) 951 (10.7) 3609 (40.5) 4221 (47.3)
180 (0.6) 27,948 (99.4) 71 (0e101) 13 (<0.01) 29 (0.1) 572 (2.0) 4507 (16.0) 13,026 (46.3) 9981 (35.5)
89 (0.9) 9920 (99.1) 67 (0e95) 1 (<0.01) 4 (<0.01) 173 (1.7) 2060 (20.6) 5310 (53.1) 2461 (24.6)
94 (0.6) 16,600 (99.4) 65 (0e102) 2 (<0.01) 19 (0.1) 499 (3.0) 5363 (32.1) 6575 (39.4) 4236 (25.4)
3 (0.9) 345 (99.1) 57 (14e94) 1 (0.3) 6 (1.7) 51 (14.7) 139 (39.9) 118 (33.9) 33 (9.5)
53 (0.3) 18,878 (99.7) 62 (0e98) 79 (0.4) 433 (2.3) 1712 (9.0) 5840 (30.8) 7464 (39.4) 3403 (18.0)
23 (0.4) 6349 (99.6) 58 (19e100) e 188 (3.0) 1536 (24.1) 1654 (26.0) 1789 (28.1) 1205 (18.9)
19 (0.3) 5694 (99.7) 40 (12e95) 2 (<0.01) 414 (6.9) 3486 (58.3) 1407 (23.5) 518 (8.7) 156 (2.6)
2 (0.4) 556 (99.6) 69 (2e95) 1 (0.2) 11 (2.0) 36 (6.5) 117 (21.0) 193 (34.6) 200 (35.8)
1 (0.7) 141 (99.3) 63 (0e95) 3 (2.1) 4 (2.8) 24 (16.9) 33 (23.2) 31 (21.8) 47 (33.1)
8340 (13.9) 12,687 (21.1) 12,835 (21.3) 14,706 (24.4) 11,582 (19.3)
2285 2071 1766 1642 1152
4627 5437 6019 6499 5546
2083 1969 2074 2129 1754
2756 3165 3504 4052 3217
(16.5) (19.0) (21.0) (24.3) (19.3)
200 (57.5) 72 (20.7) 41 (11.8) 20 (5.7) 15 (4.3)
4052 4159 4039 3922 2759
1406 1304 1353 1261 1048
(22.1) (20.5) (21.2) (19.8) (16.4)
1347 (22.5) 1369 (22.9) 1279 (21.4) 1091 (21.4) 897 (15.0)
123 (22.0) 114 (20.4) 97 (17.4) 125 (22.4) 99 (17.7)
38 30 35 27 12
146 (0.2) 9858 (16.4) 27,695 (46.0) 10,845 (18.0) 6481 (10.8) 5125(8.5)
6 (0.1) 434 (4.9) 3736 (41.9) 1825 (20.5) 1202 (13.5) 1713 (19.2)
196 (0.7) 11,386 (40.5) 5886 (20.9) 1694 (6.0) 1174 (4.2) 7792 (27.7)
66 (0.7) 4553 (45.5) 657 (6.6) 489 (4.9) 93 (0.9) 4151 (41.5)
238 (1.4) 7749 (46.4) 404 (2.4) 450 (2.7) 193 (1.2) 7660 (45.9)
4 (1.1) 177 (50.9) 67 (19.3) 44 (12.6) e 56 (16.1)
101 (0.5) 4224 (22.3) 1928 (10.2) 9892 (52.3) e 2786 (14.7)
37 (0.6) 1773 (27.8) 2438 (38.3) 1260 (19.8) 414 (6.5) 450 (7.1)
265 (4.3) 4112 (68.7) 256 (4.3) 60 (1.0) 46 (0.8) 1253 (20.9)
3 (0.5) 152 (27.2) 150 (26.9) 85 (15.2) 85 (15.2) 83 (14.9)
2 (1.4) 46 (32.4) 6 (4.2) 5 (3.5) 6 (4.2) 77 (45.2)
17,189 (29.5) 6656 (11.4) 34,395 (59.1)
5373 (60.7) 701 (7.9) 2785 (31.4)
9520 (34.0) 1513 (5.4) 16,955 (60.6)
4551 (48.3) 329 (3.5) 4543 (48.2)
6535 (42.4) 333 (2.2) 8558 (55.5)
191 (55.4) 14 (4.1) 140 (40.6)
6961 (37.0) 1988 (10.6) 9865 (52.4)
2757 (43.6) 1204 (19.0) 2365 (37.4)
2335 (39.4) 141 (2.4) 3448 (58.2)
308 (59.2) 82 (15.8) 130 (25.0)
50 (55.6) 8 (8.9) 32 (35.6)
26,551 (45.6) 18,335 (31.5) 13,355 (22.9)
7156 (80.8) 400 (4.5) 1304 (14.7)
19,719 (70.5) 1116 (4.0) 7154 (25.6)
7215 (76.6) 161 (1.7) 2047 (21.7)
10,485 (68.0) 520 (3.4) 4420 (28.7)
266 (77.1) 12 (3.5) 67 (19.4)
11,258 (59.8) 2802 (14.9) 4754 (25.3)
4735 (74.8) 343 (5.4) 1248 (19.7)
3455 (58.3) 117 (2.0) 2353 (39.7)
389 (74.8) 23 (4.4) 108 (20.8)
55 (61.1) 12 (13.3) 23 (25.6)
8285 (13.8) 23,833 (39.6) 18,198 (30.3) 242 (0.4) 9592 (15.9)
85 (1.0) 1249 (14.0) 6757 (75.8) 202 (2.3) 623 (7.0)
1368 (4.9) 8352 (29.7) 16,088 (57.2) 304 (1.1) 2016 (7.2)
2137 (21.4) 4094 (40.9) 2678 (26.8) 79 (0.8) 1021 (10.2)
7426 (44.5) 5196 (31.1) 1877 (11.2) 73 (0.4) 2122 (12.7)
30 (8.6) 123 (35.3) 85 (24.4) 9 (2.6) 101 (29.0)
2164 (11.4) 3786 (20.0) 6627 (35.0) 288 (1.5) 6066 (32.0)
285 (4.5) 2245 (35.2) 2135 (33.5) 96 (1.5) 1611 (25.3)
473 (7.9) 1457 (24.4) 1107 (18.5) 58 (1.0) 2888 (48.3)
39 (7.0) 160 (28.7) 134 (24.0) 10 (1.8) 215 (38.5)
9 (6.3) 28 (19.7) 16 (11.3) 4 (2.8) 85 (59.9)
16,825 (28.0) e 704 (1.2) 3 (<0.01) 44,342 (73.7)
6094 (68.3)
27,044 (96.1) e 8535 (30.3) 4041 (14.4) 35 (0.1)
9651 (96.4) 10009 (100) 112 (1.1) e 122 (1.2)
16,103 (96.5) e 499 (3.0) 1 (<0.01) 792 (4.7)
327 (94.0) 348 (100) 56 (16.1) 1 (0.3) 4 (1.1)
16,262 (85.9) e 13,900 (73.4) 12 (0.1) 171 (0.9)
1959 (30.7) 6372 (100) 1269 (19.9) 1 (<0.01) 19 (0.3)
5832 (97.5) e 212 (3.5) e 20 (0.3)
95 (17.0) 558 (100) 73 (13.1) e 1 (0.2)
128 (90.1) e 19 (13.4) e 1 (0.7)
RT, radiation therapy; NRT, no radiation therapy.
(25.6) (23.2) (19.8) (18.4) (12.9)
481 (5.4) 46 (0.5) 6 (0.1)
(16.4) (19.3) (21.4) (23.1) (19.7)
(20.8) (19.7) (20.7) (21.3) (17.5)
(21.4) (22.0) (21.3) (20.7) (14.6)
(26.8) (21.1) (24.6) (19.0) (8.5)
145
RTe N Z 60,150
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
Second rectal cancer Yes 462 (1.3) No 35,965 (98.7) Age, median (range) 69 (4e92) 0e14 2 (<0.01) 15e29 2 (<0.01) 30e44 27 (0.1) 45e59 3950 (10.8) 60e74 26,138 (71.8) >75 6308 (17.3) Incidence year 1989e1992 3142 (8.6) 1993e1996 5710 (15.7) 1997e2000 7599 (20.9) 2001e2004 10,605 (29.1) 2005e2008 9371 (25.7) Clinical tumour stage cT0, cTis, cTA 2 (<0.01) cT1 9363 (25.8) cT2 17,612 (48.3) cT3 8104 (22.2) cT4 838 (2.3) cTX/Unknown 508 (1.4) Clinical node stage cN0 22,299 (61.4) cNþ 649 (1.8) cNX 13,399 (36.9) Clinical metastasis stage cM0 28,798 (79.2) cM1 407 (1.1) cMX 7143 (19.7) Tumour differentiation Well 7253 (19.9) Moderate 16,416 (45.1) Poor 6577 (18.1) Undifferentiated 65 (0.2) Unknown 6116 (16.8) Treatment Surgery 1035 (2.8) Radiation therapy 36427 (100) Chemotherapy 66 (0.2) Immunotherapy 2 (<0.01) Hormonal therapy 12,987 (35.4)
Bladder
146
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
cancer compared with patients with RT-naı¨ve bladder cancer (crude incidence 0.4% versus 0.6%; p Z 0.018). 3.3. Competing risk analysis Competing risk analysis was performed to correct for death as a competing event (see Fig. 1). For all first tumour locations taken together, RT increased the risk for rectal cancer as a second primary cancer (SHR: 1.72, 95% CI: 1.55e1.91). Age did not influence this relation.
Competing risks were also calculated for each primary tumour location specifically. The risk for rectal cancer was increased in patients who received RT for prostate cancer (SHR: 1.89, 95% CI: 1.66e2.16) or endometrial cancer (SHR: 1.50, 95% CI: 1.13e2.00). Yet, a protective effect of RT was observed for patients who received RT for bladder cancer (SHR: 0.67, 95% CI: 0.47e0.94). There was no interaction with gender. No effect was observed in patients who received RT for cervical cancer (SHR: 1.17, 95% CI: 0.64e2.15), ovarian cancer (SHR:
Fig. 1. Cumulative incidence of second primary rectal cancer RT after RT for (A) any pelvic cancer; (B) prostate cancer; (C) cervical cancer; (D) endometrial cancer; (E) ovarian cancer; (F) vaginal cancer and (G) bladder cancer. RT radiation therapy; NRT, no radiation therapy.
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
2.78, 95% CI: 0.86e8.78) or vaginal cancer (SHR: 0.52, 95% CI: 0.05e5.88). 3.4. Characteristics of rectal cancer and survival analysis The median age at the time of diagnosis of rectal cancer was 75 years (range, 44e97), and most patients with a Table 3 Characteristics of rectal cancer as a second primary cancer.
Age, median (range) Gender Male Female Incidence year 1989e1992 1993e1996 1997e2000 2001e2004 2005e2008 Histology adenocarcinoma mucinous carcinoma signet ring cell carcinoma Tumour differentiation well moderate poor undifferentiated not applicable/ unknown Treatment Sx Sx þ CT Sx þ RT Sx þ CRT CT RT CRT No Sx or (C)RT Clinical tumour stage cT0/cTis cT1 cT2 cT3 cT4 cTx/unknown Clinical node stage cN0 cNþ cNx/unknown Clinical metastasis stage cM0 cM1 cMx/unknown
RT for previous pelvic cancer N Z 618 (%)
NRT for previous ppelvic cancer value N Z 751 (%)
75 (45e97)
75 (44e95)
497 (80.4) 121 (19.6)
562 (74.8) 189 (25.2)
10 (3.0) 31 (9.2) 53 (15.8) 90 (26.8) 152 (45.2)
16 (3.4) 48 (10.1) 90 (19.0) 133 (28.1) 187 (39.5)
566 (91.6) 46 (7.4)
675 (89.9) 69 (9.2)
6 (1.0)
7 (0.9)
0.002 0.014
0.541
0.511
147
rectal cancer were men (77.4%). Clinical TNM stage, differentiation grade and histology were similar between the groups (Table 3). Patients who received RT for their previous pelvic cancer were less likely to undergo additional RT for the rectal cancer (15.6% versus 46.8%, respectively, p < 0.001). Subsequently, patients who were treated with RT for the primary pelvic cancer were diagnosed with higher pathological T-stage of the rectal cancer (Table 4). There were no differences in pathological nodal stage or metastasis stage for rectal cancer in patients treated with or without RT for the primary pelvic cancer. Univariable analysis showed no statistically significant difference in overall survival between patients with rectal cancer based on their history of RT: 5-year overall survival after diagnosis of rectal cancer was 33.7% (95% CI: 29.6e37.8) in patients treated with RT for their previous pelvic cancer versus 39.1% (95% CI: 35.4e42.8) in patients without RT (Fig. 2). Multivariable analysis, correcting for age, gender, incidence year, stage and treatment of the rectal cancer, confirmed this (HR: 0.94, 95% CI: 0.79e1.11).
4. Discussion 0.182 26 (5.7) 359 (78.7) 67 (14.7) 4 (0.9) 162
31 (6.4) 357 (73.8) 94 (19.4) 2 (0.4) 267
365 (59.1) 27 (4.4) 42 (6.8) 33 (5.3) 24 (3.9) 12 (1.9) 10 (1.6) 105 (17.0)
258 (34.4) 23 (3.1) 221 (29.4) 70 (9.3) 24 (3.2) 48 (6.4) 15 (2.0) 92 (12.3)
<0.001
0.797 1 (0.3) 20 (6.1) 79 (24.0) 149 (45.3) 80 (24.3) 272
2 (0.6) 23 (6.7) 77 (22.5) 167 (48.8) 73 (21.3) 394
259 (70.4) 109 (29.6) 233
274 (65.1) 147 (34.9) 314
0.113
0.881 449 (82.2) 97 (17.8) 55
521 (82.6) 110 (17.4) 105
RT, radiation therapy; NRT, no radiation therapy; Sx, surgery; CT, chemotherapy; CRT, chemoradiation.
A recent systematic review and metaanalysis showed an increased risk of rectal cancer after RT for a variety of pelvic cancers [10]. Stratification for origin of the first cancer showed that the risk for rectal cancer was increased after treatment for prostate, cervical or endometrial cancer. The present nationwide study is the largest patient cohort in the literature and analysed the association between RT as treatment for pelvic cancer and the development of rectal cancer as a second primary cancer. The current data confirms that patients who receive RT for a primary pelvic cancer are at increased risk for the development of rectal cancer as a second primary cancer. Analysis of each primary cancer location specifically showed an increased risk for patients who received RT for a primary prostate or endometrial cancer. In contrast, a small preventive effect of RT on rectal cancer development was seen for patients who received RT for bladder cancer and no influence of RT was observed for patients who received RT for the other gynaecological cancers. Little is known about the aetiology of rectal cancer in relation to a history of RT. RT effects can be divided into set and random effects according to the probability of occurrence [17]. Set effects, such as teratogenesis, occur above a threshold dose and the severity increases with increasing the dose. In contrast, the probability for a random effect such as carcinogenesis increases as the dose increases. In that case, there is no minimal threshold dose, and the severity is independent of the dose. Since the introduction of intensity-modulated radiation therapy (IMRT), and even more volumetric
148
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
Table 4 Pathological tumour stage of rectal cancer as a second primary cancer.
No (chemo)radiation for rectal cancer
No (chemo)radiation for rectal cancer
Pathological tumour stage pT1 pT2 pT3 pT4 pTx/unknown Pathological nodal stage pN0 pNþ pNx/unknown Metastasis stage M0 M1 Mx/unknown
Pathological tumour stage ypT0 ypT1 ypT2 ypT3 ypT4 ypTx/unknown Pathological nodal stage ypN0 ypNþ ypNx/unknown Metastasis stage M0 M1 Mx/unknown
RT for previous pelvic cancer N Z 382 (%)
NRT for previous pelvic cancer N Z 277 (%)
40 (11.0) 115 (31.6) 186 (51.1) 23 (6.3) 18
69 (26.4) 82 (31.4) 100 (38.3) 10 (3.8) 16
179 (61.5) 112 (38.5) 91
107 (61.1) 68 (38.9) 102
319 (93.0) 24 (7.0) 39
213 (92.6) 17 (7.4) 47
RT for previous pelvic cancer N Z 75 (%)
NRT for previous pelvic cancer N Z 291 (%)
5 (6.9) 1 (1.4) 23 (31.9) 34 (47.2) 9 (12.5) 3
11 (4.0) 30 (10.8) 87 (31.3) 142 (51.1) 8 (2.9) 13
46 (69.7) 20 (30.3) 9
150 (58.8) 105 (41.2) 36
62 (91.2) 6 (8.8) 7
254 (93.7) 17 (6.3) 20
p-value <0.001
0.937
0.858
p-value 0.001
0.106
0.427
Pathological tumour stage of patients who underwent surgery for rectal cancer, with or without (chemo)radiation. Fourteen patients underwent neoadjuvant chemotherapy without radiation and are not taken into consideration in this table. RT, radiation therapy; NRT, no radiation therapy.
IMRT, where the high-dose region is geometrically more shaped around the target volumes at a cost of a socalled low-dose bath to a larger volume of non-target
Fig. 2. KaplaneMeier curve for overall survival after diagnosis of rectal cancer. RT radiation therapy for previous pelvic cancer; NRT, no radiation therapy for previous pelvic cancer.
tissue, it was argued that these differences in RT dose distribution might increase the risk for induction of second cancers [10,18e20]. However, little data are available to support this. There are several other possible explanations for the observed discrepancies in rectal cancer risks between the primary irradiated organs. Firstly, genetic susceptibility is likely to play a role. For instance, patients with endometrial cancer with Lynch syndrome have a 50% chance of developing a subsequent colorectal cancer [21]. In addition, DNA mismatch repair (MMR) status has been associated with radio sensitivity and long-term toxicity [22]. Increase in MMR expression is related to low-dose rate sensitivity, whereas MMR deficiency is related to high-dose rate sensitivity and possibly increased risk of carcinogenesis [22]. Secondly, because of its fixed location in the pelvis, the rectum is prone for toxicity after RT, and chronic proctitis occurs in 2e20% of patients receiving RT to the pelvis [23]. Hypothetically, RT-induced late toxicity leading to inflammation could create a precancerous environment, as is observed in patients with inflammatory bowel diseases (IBD) [24e26]. Carcinogenesis of IBD-related colorectal cancer differs from sporadic
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
colorectal cancer [25,26]. Sporadic colorectal cancer typically develops via the adenoma-carcinoma sequence, characterised by a series of molecular alterations (e.g., RAS mutation or p53 mutation). In IBD-related colorectal cancer, the ‘inflammation-dysplasia-carcinoma’ sequence shows a different pattern of molecular alterations [27]. Carcinogenesis in patients with IBD is characterised by a rapid development of high-grade dysplasia. Because RT is associated with chronic proctitis, we hypothesise that the development of rectal cancer after RT might be similar to colorectal cancer in patients with IBD. If that is the case, the incidence of rectal cancer would be related to toxicity experienced by patients after RT. For example, late grade 2 toxicity (following the Radiation Therapy Oncology Group [RTOG] and the European Organization for Research and Treatment of Cancer [EORTC]. radiation morbidity scoring schema) occurs in approximately 3e7% of patients with bladder cancer, 5e20% in patients with prostate cancer, 10e20% in patients with cervical cancer and up to 22% in patients with endometrial cancer [28e33]. In combination with the findings in the present study, it could be hypothesised that this may be an explanation for the increased risk of rectal cancer in patients with prostate and endometrial cancer. Another reason of the higher incidence of rectal cancer in patients with prostate cancer might be the dose/ volume characteristics of up to 70Gy to a small part of the anterior rectal wall in prostate cancer versus 45e50Gy to much larger volumes for most other pelvic tumours. On the other hand, numbers between groups varied and observed risk differences were small. Lack of power in the smaller groups might be one of the reasons for the lack of effect of RT for cervical, vaginal and ovarian cancer. Theoretically, differences in RT techniques, that is, external beam radiation versus brachytherapy might also play a role, but this could not be confirmed in our recent review of the literature [10]. In the NCR, no data were available on RT techniques used. In the earlier years most patients in the Netherlands received pelvic RT according to the traditional ‘box’ technique. During the 1990s, 3D conformal computed tomography (CT)e based techniques were gradually introduced, enabling individualisation to volume- instead of field-based RT. However, during multivariable analysis, incidence year of the primary cancer was not a significant predictor for the development of a rectal cancer. Differences in predisposing lifestyle factors such as smoking might also play a role in the development of second cancers. Hegemann et al. [34] have studied the incidence of second cancers after treatment for prostate cancer and concluded that the lack of information on smoking is one of the major limitations in the available literature. They reported on an increased risk for second rectal cancer in patients who were treated with RT only, but no increase in second rectal cancer was found in patients treated with surgery followed by RT or surgery
149
only. The authors hypothesised that the difference in second rectal cancer incidence might be due to differences in age and lifestyle habits such as smoking. In the present study, 36,427 patients with prostate cancer who received RT and 60,150 RT-naı¨ve patients were analysed. In contrast to the study by Hegemann et al. [34], age was not a significant predictor for the incidence of second rectal cancer. Unfortunately, information on smoking habits is not available in the NCR, and therefore, a possible confounding effect could not be studied. There was no significant difference in survival after the diagnosis of rectal cancer depending on the patients’ history of RT. Clinical stage of the second primary rectal cancers was comparable and perhaps even favourable to clinical stage reported in the literature for primary rectal cancer [35,36]. Notwithstanding, survival rates were lower than those reported for the general population (34e39% versus 51e65%) [36,37]. Patients with rectal cancer in the present study were slightly older than patients with a sporadic rectal cancer described in the literature [35,36]. Moreover, the fact that these patients have a history of cancer and cancer treatment, which might have affected their fitness and outcome, could explain the difference in overall survival. Another hypothesis could be that, tumours of patients who present with a second primary cancer might have different biological behaviour and/or might be more aggressive than tumours in patients who present with a first primary cancer [38]. This is supported by the advanced T-stage of patients with a history of RT and the relative high number of ypT4 cases after reirradiation in the present study. This nationwide study provides unique long-term data on the incidence and outcomes of second primary rectal cancer after RT for primary pelvic cancer. One of the strengths of our study is that, it contains a large cohort of patients with more than 65,000 patients having more than 10 years of follow-up and 9500 patients more than 20 years of follow-up. However, the nature of the study needs to be taken into account when interpreting the data. To diminish the influence of confounding factors, multivariable analysis was conducted. However, for some of the known predisposing factors, such as genetic susceptibility and lifestyle factors, no data were available, which is considered a limitation of the study. Only few patients underwent RT for a primary ovarian or vaginal cancer, and therefore, we could not establish whether RT for the primary cancer was related to an increased risk of second rectal cancer. However, because of the low number of patients with ovarian and vaginal cancer receiving RT, a possible relation between RT and second primary rectal cancer would not be clinically relevant.
5. Conclusions Patients who received RT for pelvic cancer were at increased risk for rectal cancer as a second primary
150
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151
cancer. However, its incidence is low and analysis per primary pelvic organ specifically showed an increased risk after prostate and endometrial cancer treatment only. Patients with previous RT for bladder cancer had a decreased risk for rectal cancer. There was no difference in survival after diagnosis of rectal cancer depending on their history of RT.
[13]
[14]
[15]
Conflict of interest statement There are no conflicts of interest to declare. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
[16]
[17]
References [1] Marijnen CA, Nagtegaal ID, Klein Kranenbarg E, Hermans J, van de Velde CJ, Leer JW, et al. No downstaging after short-term preoperative radiotherapy in rectal cancer patients. J Clin Oncol 2001;19:1976e84. [2] Rombouts AJM, Hugen N, Elferink MAG, Feuth T, Poortmans PMP, Nagtegaal ID, et al. Incidence of second tumors after treatment with or without radiation for rectal cancer. Ann Oncol 2017;28:535e40. [3] Berrington de Gonzalez A, Curtis RE, Kry SF, Gilbert E, Lamart S, Berg CD, et al. Proportion of second cancers attributable to radiotherapy treatment in adults: a cohort study in the US SEER cancer registries. Lancet Oncol 2011;12:353e60. [4] Birgisson H, Pahlman L, Gunnarsson U, Glimelius B. Occurrence of second cancers in patients treated with radiotherapy for rectal cancer. J Clin Oncol 2005;23:6126e31. [5] Rombouts AJM, Huising J, Hugen N, Siesling S, Poortmans PM, Nagtegaal ID, et al. Assessment of radiotherapy-associated angiosarcoma after breast cancer treatment in a Dutch population-based study. JAMA Oncol 2019;5(2):267e9. [6] Warschkow R, Guller U, Cerny T, Schmied BM, Plasswilm L, Putora PM. Secondary malignancies after rectal cancer resection with and without radiation therapy: a propensity-adjusted, population-based SEER analysis. Radiother Oncol 2017;123:139e46. [7] Wiltink LM, Nout RA, Fiocco M, Meershoek-Klein Kranenbarg E, Jurgenliemk-Schulz IM, Jobsen JJ, et al. No increased risk of second cancer after radiotherapy in patients treated for rectal or endometrial cancer in the randomized TME, PORTEC-1, and PORTEC-2 Trials. J Clin Oncol 2015;33: 1640e6. [8] Stenmark MH, Conlon AS, Johnson S, Daignault S, Litzenberg D, Marsh R, et al. Dose to the inferior rectum is strongly associated with patient reported bowel quality of life after radiation therapy for prostate cancer. Radiother Oncol 2014; 110:291e7. [9] Harkenrider MM, Alite F, Silva SR, Small Jr W. Image-based brachytherapy for the treatment of cervical cancer. Int J Radiat Oncol Biol Phys 2015;92:921e34. [10] Rombouts AJM, Hugen N, van Beek JJP, Poortmans PMP, de Wilt JHW, Nagtegaal ID. Does pelvic radiation increase rectal cancer incidence? a systematic review and meta-analysis. Cancer Treat Rev 2018;68:136e44. [11] Glimelius B. Recurrent rectal cancer. The pre-irradiated primary tumour: can more radiotherapy be given? Colorectal Dis 2003;5: 501e3. [12] van der Meij W, Rombouts AJ, Rutten H, Bremers AJ, de Wilt JH. Treatment of locally recurrent rectal carcinoma in
[18]
[19]
[20]
[21]
[22]
[23]
[24] [25]
[26]
[27]
[28]
[29]
[30]
previously (chemo)irradiated patients: a review. Dis Colon Rectum 2016;59:148e56. Susko M, Lee J, Salama J, Thomas S, Uronis H, Hsu D, et al. The use of re-irradiation in locally recurrent, non-metastatic rectal cancer. Ann Surg Oncol 2016;23:3609e15. Guren MG, Undseth C, Rekstad BL, Braendengen M, Dueland S, Spindler KL, et al. Reirradiation of locally recurrent rectal cancer: a systematic review. Radiother Oncol 2014;113: 151e7. Valentini V, Morganti AG, Gambacorta MA, Mohiuddin M, Doglietto GB, Coco C, et al. Preoperative hyperfractionated chemoradiation for locally recurrent rectal cancer in patients previously irradiated to the pelvis: a multicentric phase II study. Int J Radiat Oncol Biol Phys 2006;64:1129e39. Rombouts AJ, Koh CE, Young JM, Masya L, Roberts R, DeLoyde K, et al. Does radiotherapy of the primary rectal cancer affect prognosis after pelvic exenteration for recurrent rectal cancer? Dis Colon Rectum 2015;58:65e73. ICRP. ICoRP. Publication 84. Pregnancy and medical irradiation. Ann ICRP 2000;30:1e43. Kry SF, Salehpour M, Followill DS, Stovall M, Kuban DA, White RA, et al. The calculated risk of fatal secondary malignancies from intensity-modulated radiation therapy. Int J Radiat Oncol Biol Phys 2005;62:1195e203. Berrington de Gonzalez A, Wong J, Kleinerman R, Kim C, Morton L, Bekelman JE. Risk of second cancers according to radiation therapy technique and modality in prostate cancer survivors. Int J Radiat Oncol Biol Phys 2015;91:295e302. Lonn S, Gilbert ES, Ron E, Smith SA, Stovall M, Curtis RE. Comparison of second cancer risks from brachytherapy and external beam therapy after uterine corpus cancer. Cancer Epidemiol Biomark Prev 2010;19:464e74. Win AK, Lindor NM, Winship I, Tucker KM, Buchanan DD, Young JP, et al. Risks of colorectal and other cancers after endometrial cancer for women with Lynch syndrome. J Natl Cancer Inst 2013;105:274e9. Martin LM, Marples B, Coffey M, Lawler M, Lynch TH, Hollywood D, et al. DNA mismatch repair and the DNA damage response to ionizing radiation: making sense of apparently conflicting data. Cancer Treat Rev 2010;36:518e27. Wu XR, Liu XL, Katz S, Shen B. Pathogenesis, diagnosis, and management of ulcerative proctitis, chronic radiation proctopathy, and diversion proctitis. Inflamm Bowel Dis 2015;21: 703e15. Sjodahl RI, Myrelid P, Soderholm JD. Anal and rectal cancer in Crohn’s disease. Colorectal Dis 2003;5:490e5. Dyson JK, Rutter MD. Colorectal cancer in inflammatory bowel disease: what is the real magnitude of the risk? World J Gastroenterol 2012;18:3839e48. Baker KT, Salk JJ, Brentnall TA, Risques RA. Precancer in ulcerative colitis: the role of the field effect and its clinical implications. Carcinogenesis 2018;39:11e20. Keller DS, Windsor A, Cohen R, Chand M. Colorectal cancer in inflammatory bowel disease: review of the evidence. Tech Coloproctol 2019;23:3e13. Aluwini S, van Rooij PH, Kirkels WJ, Boormans JL, KolkmanDeurloo IK, Wijnmaalen A. Bladder function preservation with brachytherapy, external beam radiation therapy, and limited surgery in bladder cancer patients: long-term results [corrected]. Int J Radiat Oncol Biol Phys 2014;88:611e7. Chen CC, Wang L, Lu CH, Lin JC, Jan JS. Comparison of clinical outcomes and toxicity in endometrial cancer patients treated with adjuvant intensity-modulated radiation therapy or conventional radiotherapy. J Formos Med Assoc 2014;113: 949e55. Kishan AU, Kupelian PA. Late rectal toxicity after low-dose-rate brachytherapy: incidence, predictors, and management of side effects. Brachytherapy 2015;14:148e59.
A.J.M. Rombouts et al. / European Journal of Cancer 124 (2020) 142e151 [31] Mohammed N, Kestin L, Ghilezan M, Krauss D, Vicini F, Brabbins D, et al. Comparison of acute and late toxicities for three modern high-dose radiation treatment techniques for localized prostate cancer. Int J Radiat Oncol Biol Phys 2012;82: 204e12. [32] Chopra S, Dora T, Engineer R, Mechanery S, Agarwal P, Kannan S, et al. Late rectal toxicity after image-based high-doserate interstitial brachytherapy for postoperative recurrent and/or residual cervical cancers: EQD2 predictors for Grade >/ZII toxicity. Brachytherapy 2015;14:881e8. [33] Mattes MD, Kan CC, Dalbagni G, Zelefsky MJ, Kollmeier MA. External beam radiation therapy for small cell carcinoma of the urinary bladder. Pract Radiat Oncol 2015;5:e17e22. [34] Hegemann NS, Schlesinger-Raab A, Ganswindt U, Horl C, Combs SE, Holzel D, et al. Risk of second cancer following radiotherapy for prostate cancer: a population-based analysis. Radiat Oncol 2017;12:2.
151
[35] Benitez Majano S, Di Girolamo C, Rachet B, Maringe C, Guren MG, Glimelius B, et al. Surgical treatment and survival from colorectal cancer in Denmark, England, Norway, and Sweden: a population-based study. Lancet Oncol 2019;20:74e87. [36] Brouwer NPM, Bos A, Lemmens V, Tanis PJ, Hugen N, Nagtegaal ID, et al. An overview of 25 years of incidence, treatment and outcome of colorectal cancer patients. Int J Cancer 2018;143:2758e66. [37] Peeters KC, Marijnen CA, Nagtegaal ID, Kranenbarg EK, Putter H, Wiggers T, et al. The TME trial after a median followup of 6 years: increased local control but no survival benefit in irradiated patients with resectable rectal carcinoma. Ann Surg 2007;246:693e701. [38] VanderWalde AM, Hurria A. Second malignancies among elderly survivors of cancer. The Oncologist 2011;16:1572e81.